Microvolume laser scanning cytometry (MLSC) is a method for analyzing the expression of biological markers in a biological fluid. See, e.g., U.S. Pat. Nos. 5,547,849 and 5,556,764; Dietz et al., Cytometry 23:177–186 (1996); U.S. Provisional Application 60/144,798, filed Jul. 21, 1999, each of which is incorporated herein by reference. A sample, such as blood, is incubated in a capillary with one or more fluorescently-labeled probes that specifically binds to particular biological markers, such as membrane proteins displayed on the surface of a blood cell. The sample is then analyzed by a MLSC instrument, which scans excitation light from a laser over the sample along one axis of the capillary, while the capillary itself is moved in an orthogonal axis by an automated stage. Fluorescent probes in the sample emit Stokes-shifted light in response to the excitation light, and this light is collected by the cytometer and used to form an image of the sample. In such images, the cells or particles that bind to the fluorescent probes can be identified and quantitated by image analysis algorithms. The resulting information on the expression of biological markers in the sample can be used for diagnostic and prognostic medical purposes.
Current laser scanning cytometers are based on the flying spot confocal laser scanner. These systems scan a laser excitation light spot in one dimension across the sample using a rotating or reciprocating mirror, such as a mirror mounted on a galvanometer. The sample is translated in a direction orthogonal to the scan direction. The collimated excitation laser light follows an epi-illumination path through the microscope objective and is focused on the sample and the mirror scan center is imaged upon the entrance pupil of the microscope objective. Emitted light from the sample is then collected by the microscope objective, and re-traces the excitation light path back to the scanning mirror where it is descanned. The emitted light passes through a dichroic filter and a long-pass filter to separate out reflected excitation light, and is then focussed onto an optical detector through an aperture. The aperture serves as a spatial filter, and reduces the amount of out-of-focus light that is introduced into the detector. The wider the aperture, the greater the depth of focus of the system. The detector generates a signal that is proportional to the intensity of the incident light. Thus, as the laser scans the sample, an image is assembled pixel-by-pixel. This optical architecture is typically referred to as confocal fluorescence detection.
In order to detect multiple fluorescence probes, the laser scanning cytometry system can also include dichroic filters that separate the emitted light into its component wavelengths. Each distinct wavelength is imaged onto a separate detector through a separate aperture. In this way, an image of the sample is assembled pixel-by-pixel for each emitted wavelength. The individual images are termed channels, and the final multi-color image is obtained by merging the individual channels.
The use of confocal 4-channel fluorescence detection for MLSC is illustrated schematically in FIG. 1, and is described in the United States Provisional Patent Application entitled “Improved System for Microvolume Laser Scanning Cytometry”, filed Jul. 21, 1999, incorporated herein by reference in its entirety. In this embodiment, the light from a laser is scanned over a capillary array 8 wherein each capillary contains a sample that contains fluorescently-labeled species. Specifically, collimated excitation light is provided by a He-Ne laser 10. The collimated laser light is deflected by an excitation dichroic filter 12. Upon reflection, the light is incident on a galvanometer-driven scan mirror 14. The scan mirror can be rapidly oscillated over a fixed range of angles by the galvanometer e.g. +/−2.5 degrees. The scanning mirror reflects the incident light into two relay lenses (relay lens 1 (16) and relay lens 2 (18)) that image the scan mirror onto the entrance pupil of the microscope objective 20. This optical configuration converts a specific scanned angle at the mirror to a specific field position at the focus of the microscope objective 20. The angular sweep is typically chosen to result in a 1 mm scan width at the objective's focus. The relationship between the scan angle and the field position is essentially linear in this configuration and over this range of angles. Furthermore the microscope objective 20 focuses the incoming collimated beam to a spot at the objective's focus plane. The spot diameter, which sets the optical resolution, is determined by the diameter of the collimated beam and the focal length of the objective. Fluorescence samples placed in a capillary array 8 in the path of the swept excitation beam emit stokes-shifted light. This light is collected by the objective 20 and collimated. This collimated light emerges from the two relay lenses (16 and 18) still collimated and impinges upon the scan mirror 14 which reflects and descans it. The stokes-shifted light then passes through an excitation dichroic filter 12 (most excitation light reflected within the optics to this point is reflected by this dichroic filter) and then through long pass filter 1 (22) that further serves to filter out any reflected excitation light. Fluorescence dichroic filter 1 (24) then divides the two bluest fluorescence colors from the two reddest. The two bluest colors are then focussed through focusing lens 1 (26) onto aperture 1 (28) in order to significantly reduce any out-of-focus fluorescence signal. After passing though the aperture, fluorescence dichroic 2 (30) further separates the individual blue colors from one another. The individual blue colors are then parsed to two separate photomultipliers 1 (32) and 2 (34). After being divided from the two bluest colors by fluorescence dichroic 1 (24), the two reddest colors are passed through long pass filter 2 (36) and reflected off a mirror (38) through focusing lens 2 (40) onto aperture 42. After passing through aperture 2, the reddest colors are separated from one another by fluorescence dichroic 3 (44). The individual red colors are then parsed to photomultipliers 3 (46) and 4 (48). In this way, four separate fluorescence signals can be simultaneously transmitted from the sample held in the capillary to individual photomultiplier light detectors (PMT1-4). Each photomultiplier generates an electronic current in response to the incoming fluorescence photon flux. These individual currents are converted to separate voltages by one or more preamplifiers in the detection electronics. The voltages are sampled at regular intervals by an analog to digitial converter in order to determine pixel intensity values for the scanned image.
Other ways are known in the art for obtaining multi-channel information in the microscopy context. For example, it is known in the art to use fluorophores that emit light with overlapping emission spectra but with different time constants of emission. Time-resolved microscopy systems typically use very fast laser pulses and high speed detection circuitry to resolve, in the time-domain, the nanosecond-scale time signatures of the fluorophores. Alternatively, the measurement can be accomplished in the frequency domain with amplitude-modulated laser sources and detection circuitry that measures phase shift and modulation amplitudes. Both of these techniques add significant complexity to the fluorescent measurement system.
Typical MLSC instruments use photomultiplier tubes (PMTs) as light detectors. PMTs are cost-effective and have high data read-out rates which allow the sample to be scanned swiftly. However, a major drawback of the PMT is its low quantum efficiency. For example, in the red to near infrared region of the electromagnetic spectrum, PMTs have quantum efficiencies of less than 10%, i.e., less than one photon in ten that impacts the PMT is actually detected.
In order to have high sensitivity and high measurement speed, it is desirable to use high power laser sources. For each fluorophore, there is a proportional relationship between the intensity of the excitation illumination and the intensity of the emitted light. This proportionality only applies up to the point at which the fluorophore is saturated. At this point, the ground state of the fluorophore is essentially depleted, and all fluorophores exist in the excited electronic configuration. Increasing laser power beyond the saturation point does not increase the intensity of the emitted light. This effect is especially pronounced with fluorophores that have long fluorescent lifetimes, such as inorganic fluorophores, and for quantum dot nanocrystals. These molecules saturate at relatively low power densities because of their long time constants of fluorescence emission. Other undesired processes can occur at higher laser power, including photodestruction and intersystem crossing. In many applications, it is desirable to operate at power densities somewhat below saturation.
In an attempt to increase the speed with which confocal images can be acquired, microscope systems have been developed in which a continuous line of laser excitation light is scanned across the sample, rather than a single spot. The line of emitted light that is produced by the sample in response to the excitation is imaged onto a slit shaped aperture. Since light is distributed over a line of pixels speed limitations due to fluorophore photo-physics are avoided. However, the depth of field, or change in lateral resolution with focus depth, of the line scanner is inversely proportional to the numerical aperture of the objective. MLSC applications require a large depth of field to accurately image cells in a deep blood suspension. High sensitivity and speed require a high numerical aperture lens but would result in a prohibitively small depth of field. This tradeoff ultimately results in limited speed and sensitivity.
Given the limitations using a PMT as a light detector, much research is currently directed towards developing higher efficiency detectors that will allow rapid image acquisition at below-saturation power densities. One such light detector is the charge coupled device (CCD). See, e.g., G. Holst, CCD Arrays, Camera and Displays, 2d Ed., JCD Publishing and SPIE Optical Engineering Press 1998. The CCD consists of an interlinked array of sensitive photodetectors, each of which can have a quantum efficiency of greater than 80%. Despite their high efficiency, CCDs are not ideal for use in the MLSC context. One reason is that the CCD is usually employed as an imaging device in which the entire field of view is excited and the CCD captures all of the emitted light in the field of view. Used in this way the depth of field and sensitivity are coupled just as in a line scanner. Furthermore, the full field illumination and collection means a substantial amount out of focus light is excited and received by the CCD detector. Even if a CCD is used in non-imaging mode in combination with a scanned laser spot and a pinhole aperture—much in the same way as a PMT is used—additional problems are encountered. Firstly, when replacing PMTs with CCDs for simultaneous multi-channel image acquisition, a separate CCD is required for each channel. The cost of providing a separate high-efficiency CCD for each channel adds greatly to the cost of the instrument. Moreover, CCDs take significantly longer to read out than PMTs, thereby placing a significant limitation on the speed with which data can be acquired.
The present invention is directed towards optical architectures for spectroscopy which can acquire data with greater speed than prior art systems. The instruments of the invention can also be used for time resolved measurements of fluorophore emission. The methods and instruments of the invention are useful in any application where spectroscopic data from a sample is required. In preferred embodiments, the methods and instruments of the invention are used for MLSC.
The invention uses CCDs in which binning is used to subdivide a single CCD into pixel groupings that collect data simultaneously from a number of different regions of a sample. Preferred embodiments of the invention use multiple laser excitation spots in combination with CCD light detectors. In some embodiments, the individual bins are further subdivided to provide spectral information for each region of the sample. The pixel intensity values for each bin are assembled by computer to give seamless images of the sample in each channel.